Question: Configuring Pulsar Failed
0
yuedeji • 0 wrote:
We are trying to move Galaxy jobs run on a remote HPC cluster and return the results back. Pulsar seems to be a good fit. Our current solution is moving Galaxy jobs to a Pulsar server, and then let Pulsar push jobs to a HPC cluster.
Galaxy Server --1--> Pulsar Server --2--> HPC Cluster
However, we failed to make it work at the step 1. We successfully installed both Galaxy and Pulsar, but we failed to make the connect. The configurations are listed below. Could someone tell us where did we configure incorrectly? Really appreciate it!
On Galaxy side, the job_conf.xml file is
<?xml version="1.0"?>
<job_conf>
<plugins>
<plugin id="local" type="runner" load="galaxy.jobs.runners.local:LocalJobRunner"/>
<plugin id="pulsar" type="runner" load="galaxy.jobs.runners.pulsar:PulsarRESTJobRunner"/>
</plugins>
<handlers>
<handler id="main"/>
</handlers>
<destinations default="local">
<destination id="local" runner="local"/>
<destination id="remote_cluster" runner="pulsar">
<param id="url">http://xxx.xxx.xxx.xxx:8913/</param>
<param id="private_token">abc</param>
</destination>
</destinations>
<tools>
<tool id="msconvert" destination="remote_cluster" />
<tool id="bed_to_bigBed" destination="remote_cluster" />
<tool id="proteinpilot" destination="remote_cluster" />
</tools>
</job_conf>
On Pulsar side, the server.ini file is [server:main] use = egg:Paste#http port = 8913 host = localhost ssl_pem = host.pem
app.yml is
private_key: abc